Asset Production
Generate all the visual elements, audio, and videos for your project using AI tools.
🖼️ Generating Product Images
Images are the visual foundation of your video. Let's generate high-quality product images in different settings and angles using Midjourney and FLUX.
Images Needed (Typical for 30s)
| Scene | Image type | Base prompt |
|---|---|---|
| Hero shot | Featured Product, Premium Background | "product photography, [product], dramatic lighting, dark background" |
| Context | Product in Everyday Use | "person using [product], lifestyle photography, natural light" |
| Detail | Close-up of details and textures | "macro photography, [product detail], sharp focus" |
| Scenario | Background/environment for replacement | "[scene description], wide shot, cinematic, --ar 16:9" |
Midjourney vs. FLUX: When to Use Each
- • Midjourney: Best for aesthetic, artistic, and product images with dramatic lighting
- • FLUX: Best for images with legible text, specific compositions, and photorealism
- • Tip: Generate 3-4 variations of each image and choose the best one
Image production workflow
- Review your storyboard and identify each image needed
- Write prompts based on the script’s visual descriptions
- Generate 3-4 variations for each scene in Midjourney
- Select the best ones and upscale them (maximum resolution)
- If needed, use FLUX for images that need text
- Save all the images in the folder
/assets/imagens/ - Name them in an organized way:
cena01-hero.png,cena02-contexto.png
🎥 Generating Video Clips
AI-generated video clips add dynamism to the project. Use Kling 3.0 for realistic videos and Runway for more artistic effects.
AI video strategies for the project
Image-to-Video
Use the images generated in Midjourney as a starting point. Upload them to Kling 3.0 and animate the scene with subtle movements: slow motion, zoom, object movement.
Text-to-Video
For more dynamic scenes, describe the action in text and let AI generate the full video. Ideal for transition scenes and B-roll.
Tips for better results
| Tool | Ideal duration | Key Tip |
|---|---|---|
| Kling 3.0 | 5-10 seconds | Describe camera movements in the prompt (pan, zoom, orbit) |
| Runway Gen-4.5 | 4–8 seconds | Use a reference image for visual consistency |
| Pika | 3-5 seconds | Great for short, eye-catching creative effects |
Don't try to generate everything as video
For a 30-second video, 2-3 AI video clips are enough. The rest can be made up of still images with camera movements (zoom, pan) added in CapCut. Mixing animated images with video clips is the most efficient strategy.
🎙️ Recording the Voiceover with ElevenLabs v3
Narration brings your video to life. ElevenLabs v3 (ElevenMusic) (ElevenMusic) generates incredibly natural Brazilian Portuguese voices with fine control over emotion, pacing, and intonation.
Step-by-step narration guide
- Go to the ElevenLabs > Text to Speech
- Select a Brazilian Portuguese voice (test 3-4 different voices)
- Paste the narration text from your script
- Adjust the settings:
- Stability: 50-70% (more natural variation)
- Clarity: 75–85% (clear voice, but not robotic)
- Style: Adjust to match the video’s tone
- Generate the narration and listen carefully
- Adjust words or phrases that sound strange
- Export as MP3 or WAV (high quality)
Tip: Pauses and emphasis
Use strategic punctuation in the text to control the narration: Commas = short pauses, Points = long pauses, UPPERCASE = emphasis, ... = dramatic pause. Example: "What if I told you... this costs LESS than 50 reais?"
🎵 Creating Background Music with Suno v5
Music sets the video’s emotional atmosphere. Suno v5.5 generates original music that fits your project perfectly.
Background music prompt
Estilo: [energético/calmo/inspirador/corporativo] BPM: [80-100 para calmo, 120-140 para energético] Instrumentos: [piano, synth, beats eletrônicos, violão] Mood: [motivacional, sofisticado, divertido, dramático] Duração: [30s ou 60s] Sem vocais (instrumental only)
Music styles by video type
| Video type | Musical style | BPM |
|---|---|---|
| Tech/Gadget Product | Electronic, futuristic, strong beats | 120-130 |
| Beauty Product | Soft, elegant, piano + synth pad | 90-110 |
| Food/Drink | Cheerful, acoustic, guitar + light percussion | 100-120 |
| Corporate/SaaS | Motivational, light orchestral music + beats | 110-125 |
| Fashion/Lifestyle | Lo-fi, chill, mellow beats | 85-100 |
Background music volume
Background music should be -15dB to -20dB below the narration. In CapCut, set the music track volume to 15-25% of the total. The music should be felt, not consciously heard—it complements the narration and never competes with it.
📂 Organizing the Assets
With all the assets generated, organize them clearly before moving on to assembly.
Recommended folder structure
projeto-final/ ├── imagens/ │ ├── cena01-hero-produto.png │ ├── cena02-contexto-uso.png │ ├── cena03-detalhe.png │ └── cena04-cenario-fundo.png ├── videos/ │ ├── cena01-intro-animado.mp4 │ ├── cena03-transicao.mp4 │ └── cena04-montagem.mp4 ├── audio/ │ ├── narracao-completa.mp3 │ └── musica-fundo.mp3 ├── exports/ │ └── (vídeos finalizados aqui) └── roteiro.txt
Asset Quality Checklist
✅ Lesson Checklist
- I generated all the necessary images (Midjourney/FLUX)
- I generated the video clips (Kling 3.0/Runway)
- I recorded the full voiceover (ElevenLabs v3)
- Created the background music (Suno v5)
- I checked the quality of all assets
- I organized all files into correctly named folders